Why contextual preference reversals maximize expected value

Contextual preference reversals occur when a preference for one choice over another is reversed by the addition of further choices. It has been argued that the occurrence of such preference reversals in human choice are a violation of axioms of rational choice, and so are not consistent with utility maximization. In contrast, we demonstrate here that for an agent that obtains noisy estimates of expected value and the order of feature values, the rational, expected value maximizing choice between existing options is influenced by the addition of new choices, and does give rise to apparent preference reversals. We furthermore show, through computational simulations that derive the implications of the rational integration of the agent's observations, that utility maximizing choice exhibits a range of contextual preference reversal types|including attraction, compromise, similarity and phantom effects. These preference reversal types have played a key role in the development of models of human choice. We conclude that experiments showing that people exhibit contextual preference reversals are not evidence that they are irrational; they are however evidence that they make use of the available information in the presence of uncertainty to maximize utility.